An up‑to‑date, stream‑ready mirror of the official OpenLibrary public data dumps, converted to snappy‑compressed Parquet.
openlibrary/
├── authors/ # authors_0.parquet, authors_1.parquet, … (≈ 1-2 GB each)
├── editions/ # editions_0.parquet, editions_1.parquet, …
└── works/ # works_0.parquet, works_1.parquet, …
from datasets import… See the full description on the dataset page:
https://huggingface.co/datasets/sayshara/openlibrary.
